Two-phase cold plate and spray type charging pile intelligent heat dissipation system

Abstract

The present application relates to the technical field of charging pile heat dissipation, and discloses a charging pile intelligent heat dissipation system based on a two-phase cold plate and a spraying type. The system comprises a heat dissipation area division unit, a data acquisition unit, a heat dissipation mode analysis unit, an environmental heat dissipation area determination unit, a heat dissipation weight construction unit, a heat dissipation sequence generation unit and a driving execution unit. The heat dissipation area division unit divides the charging pile into multiple heat dissipation adjustment areas; the data acquisition unit acquires temperature distribution data of the surface contact area of the shell in real time; the heat dissipation mode analysis unit identifies the target heat dissipation area accordingly, and divides the coverage area thereof into a dominant heat dissipation area and an auxiliary heat dissipation area; the environmental heat dissipation area determination unit determines the environmental heat dissipation area in the uncovered area, and the three areas constitute a heat dissipation processing set. The heat dissipation weight construction unit calculates the heat dissipation adjustment difference between adjacent areas and constructs a weight map, and the heat dissipation sequence generation unit generates a heat dissipation adjustment sequence accordingly; the driving execution unit drives the two-phase cold plate and the spraying module to dissipate heat according to the sequence, thereby achieving targeted heat dissipation.

Description

Technical Field

[0001] This invention relates to the field of charging pile heat dissipation technology, specifically to an intelligent heat dissipation system for charging piles based on a two-phase cold plate and a spray system. Background Technology

[0002] With the rapid development of the new energy vehicle industry, charging piles, as a key infrastructure for energy replenishment, are experiencing increasingly higher usage frequency and power demands. In high-power charging scenarios, the electronic components inside the charging pile continuously generate a large amount of heat. If this heat cannot be dissipated in time, the internal temperature will rise sharply, affecting the operational stability and lifespan of the electronic components. Currently, common heat dissipation methods for charging piles on the market mainly include natural cooling, forced air cooling, and liquid cooling alone.

[0003] Natural cooling relies on the natural heat exchange between the charging pile's casing and the external environment, requiring no additional power and thus having a lower cost. However, its cooling efficiency is greatly affected by the ambient temperature, making it difficult to meet cooling requirements in high-temperature environments or high-power charging conditions. Forced air cooling uses fans to drive airflow, accelerating the removal of heat from the charging pile's interior. Compared to natural cooling, it offers improved cooling efficiency, but the fans generate noise during operation, and dust easily accumulates inside them after prolonged use, further reducing cooling efficiency and potentially causing equipment malfunctions. Liquid cooling typically involves direct contact between a cold plate and the heat-generating components for heat exchange. While its cooling effect is superior to air cooling, traditional liquid cooling systems often employ a holistic cooling design, providing overall cooling for the entire charging pile and failing to allow for targeted adjustments based on the heat generation in different areas.

[0004] In practical applications, the heat intensity of electronic components in different locations within a charging pile varies significantly. For example, the heat density of components such as power modules and control units is much higher than that of other auxiliary components. When adopting an integrated heat dissipation design, to ensure effective heat dissipation in areas with high heat intensity, the operating power of the entire heat dissipation system needs to be increased. This not only wastes energy but may also lead to excessive heat dissipation in areas with low heat intensity, resulting in energy loss and potentially affecting the normal operation of some components due to excessively low local temperatures. Furthermore, traditional heat dissipation systems lack real-time monitoring of the surface temperature distribution of the charging pile's outer casing, making it impossible to promptly grasp the internal heat distribution and adjust the heat dissipation strategy according to the actual heat dissipation conditions. This results in poor operational flexibility and adaptability of the heat dissipation system, failing to meet the efficient heat dissipation requirements of charging piles under different operating conditions. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent heat dissipation system for charging piles based on two-phase cold plates and spraying, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides an intelligent heat dissipation system for charging piles based on a two-phase cold plate and a spray system, the system comprising:

[0007] The heat dissipation area division unit is used to divide the interior of the charging pile into multiple heat dissipation adjustment zones;

[0008] The data acquisition unit is used to collect real-time temperature distribution data of the contact area on the surface of the charging pile casing.

[0009] The heat dissipation mode analysis unit is used to identify the target heat dissipation area based on the temperature distribution data of the contact area, and to divide the heat dissipation adjustment area covered by the target heat dissipation area into the main heat dissipation area and the auxiliary heat dissipation area.

[0010] An environmental heat dissipation zone determination unit is used to determine an environmental heat dissipation zone in a heat dissipation adjustment zone not covered by the target heat dissipation zone. The main heat dissipation zone, the auxiliary heat dissipation zone and the environmental heat dissipation zone constitute a heat dissipation treatment set.

[0011] A heat dissipation weight construction unit is used to calculate the heat dissipation adjustment difference between any two adjacent heat dissipation adjustment zones in the heat dissipation treatment set, and construct a heat dissipation weight graph with the heat dissipation adjustment zone as the vertex and the heat dissipation adjustment difference as the edge.

[0012] A heat dissipation sequence generation unit is used to generate a heat dissipation adjustment sequence based on the heat dissipation weight map;

[0013] The drive execution unit is used to drive the two-phase cold plate module and the spray module to perform heat dissipation operations according to the heat dissipation adjustment sequence.

[0014] Preferably, the heat dissipation mode analysis unit includes:

[0015] The contact area identification module is used to build a standard contact area model based on historical temperature data;

[0016] The area matching module is used to match the real-time collected temperature distribution data of the contact area with the standard contact area model to identify the boundary of the target heat dissipation area under the current operating state of the charging pile.

[0017] The region mapping module is used to map the boundary of the target heat dissipation area to the spatial coordinates of the heat dissipation adjustment area.

[0018] Preferably, the contact area identification module is specifically used for:

[0019] Construct a three-dimensional temperature distribution matrix on the surface of the charging pile casing;

[0020] The K-nearest neighbor algorithm is used to perform cluster analysis on the three-dimensional temperature distribution matrix to generate a set of regions with similar temperature characteristics;

[0021] Effective contact areas are selected from the set of temperature-characteristic similar regions based on the temperature change gradient.

[0022] Preferably, the heat dissipation mode analysis unit further includes:

[0023] The heat source analysis module is used to extract thermal imaging feature data within the target heat dissipation area;

[0024] The dominant region determination module marks the heat dissipation adjustment area corresponding to the thermal imaging feature data as the dominant heat dissipation area based on the heat flux density threshold.

[0025] The auxiliary area determination module marks the heat dissipation adjustment area adjacent to the main heat dissipation area as the auxiliary heat dissipation area based on the heat conduction gradient.

[0026] Preferably, the environmental heat dissipation zone determination unit is specifically used for:

[0027] Establish a spatial relationship map of the heat dissipation regulation zones;

[0028] Density peak clustering algorithm is used to identify the core influence area of ​​the dominant heat dissipation zone;

[0029] The range of the environmental heat dissipation zone is determined in the non-target heat dissipation zone based on the thermal diffusion attenuation model.

[0030] Preferably, the heat dissipation weighting construction unit includes:

[0031] The parameter calculation module is used to obtain the heat load parameters and heat dissipation efficiency parameters of each heat dissipation adjustment zone;

[0032] The difference calculation module uses a fuzzy logic algorithm to process the difference between the heat load parameters and the heat dissipation efficiency parameters of adjacent heat dissipation adjustment zones, and outputs the difference in heat dissipation adjustment.

[0033] Preferably, the heat dissipation sequence generation unit is specifically used for:

[0034] Convert the heat dissipation weight graph into an adjacency matrix;

[0035] The optimal path sequence of the adjacency matrix is ​​obtained based on the Hungarian algorithm;

[0036] The optimal path sequence is optimized based on the working state constraints of the heat dissipation adjustment zone.

[0037] Preferably, the drive execution unit includes:

[0038] The timing control module is used to parse the execution node parameters in the heat dissipation adjustment sequence;

[0039] The power prediction module uses a long short-term memory network model to process historical heat dissipation power data and generate phase change triggering parameters for the two-phase cold plate module and flow control parameters for the spray module.

[0040] The collaborative drive module synchronously adjusts the refrigerant flow rate and spray intensity based on the phase change trigger parameters and flow control parameters.

[0041] Preferably, the system further includes:

[0042] An abnormal heat dissipation detection unit is used to monitor the temperature change rate of the heat dissipation processing assembly;

[0043] The pattern matching unit compares the real-time temperature change rate curve with a preset abnormal heat dissipation pattern library;

[0044] The parameter correction unit updates the edge weight values ​​of the heat dissipation weight graph when the matching degree exceeds a set threshold.

[0045] Preferably, the pattern matching unit is specifically used for:

[0046] Construct a time-series feature vector of the temperature change rate;

[0047] The similarity between the time series feature vector and the abnormal heat dissipation pattern template is calculated using a dynamic time warping algorithm.

[0048] The abnormal heat dissipation pattern type with the highest matching degree is determined based on the similarity ranking results.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] By setting up heat dissipation zone division units, the interior of the charging pile is divided into multiple heat dissipation adjustment zones, breaking the limitations of traditional integrated heat dissipation design and providing a foundation for targeted heat dissipation. This zoned design can accurately correspond to the areas where electronic components with different heat intensities are located inside the charging pile, allowing subsequent heat dissipation operations to be carried out according to the actual heat dissipation conditions of each area, avoiding energy waste caused by indiscriminate heat dissipation of the entire charging pile interior.

[0051] The data acquisition unit can collect real-time temperature distribution data of the contact area on the surface of the charging pile's casing. The surface temperature distribution is directly related to the heating status of internal components. By collecting this data, the dynamic distribution of heat inside the charging pile can be grasped in a timely and accurate manner, providing a reliable source of information for subsequent identification of target heat dissipation areas. Compared to traditional heat dissipation systems that lack real-time monitoring, this unit allows the heat dissipation system to dynamically respond to changes in internal heat generation, no longer relying on a preset fixed heat dissipation mode, thus improving the heat dissipation system's ability to perceive actual operating conditions.

[0052] The heat dissipation mode analysis unit identifies the target heat dissipation area based on the temperature distribution data of the contact area and divides the heat dissipation adjustment area covered by the target heat dissipation area into a primary heat dissipation area and an auxiliary heat dissipation area. This division further refines the heat dissipation requirements. The primary heat dissipation area corresponds to the area with high internal heat intensity, while the auxiliary heat dissipation area corresponds to the area with relatively low heat intensity but still requires enhanced heat dissipation. By clarifying the heat dissipation priority of different areas, subsequent heat dissipation operations can prioritize the heat dissipation needs of areas with high heat intensity while also taking into account the auxiliary heat dissipation areas, avoiding the problem of insufficient or excessive heat dissipation in some areas due to unreasonable allocation of heat dissipation resources.

[0053] The environmental heat dissipation zone determination unit identifies the environmental heat dissipation zone within the heat dissipation adjustment area not covered by the target heat dissipation area. This allows the primary heat dissipation zone, auxiliary heat dissipation zone, and environmental heat dissipation zone to collectively form a heat dissipation treatment system, achieving comprehensive coverage of all areas inside the charging pile. The environmental heat dissipation zone corresponds to areas with extremely low internal heat intensity that can be met by natural heat dissipation, eliminating the need for additional activation of the two-phase cold plate module and spray module for active heat dissipation. This further reduces energy consumption and improves the overall energy efficiency of the heat dissipation system.

[0054] The heat dissipation weight construction unit constructs a heat dissipation weight graph by calculating the heat dissipation adjustment difference between any two adjacent heat dissipation adjustment zones in the heat dissipation processing set. This graph has heat dissipation adjustment zones as vertices and heat dissipation adjustment difference as edges. This weight graph clearly reflects the differences in heat dissipation requirements between adjacent areas. Based on this weight graph, the heat dissipation sequence generation unit can generate a scientifically sound heat dissipation adjustment sequence, ensuring that heat dissipation operations are carried out in an orderly manner according to the differences in heat dissipation requirements between areas. This avoids the disorder and randomness of heat dissipation operations, enabling the optimal allocation of heat dissipation resources based on the differences in regional requirements.

[0055] The drive unit drives the two-phase cold plate module and the spray module to perform heat dissipation operations according to the heat dissipation adjustment sequence. The two-phase cold plate module has a high-efficiency heat exchange capacity, which can quickly remove a large amount of heat from the main heat dissipation area and the auxiliary heat dissipation area, while the spray module can further enhance the heat dissipation effect by spraying coolant. The synergistic work of the two heat dissipation modules can give full play to their respective heat dissipation advantages and improve the overall heat dissipation efficiency. At the same time, performing heat dissipation operations according to the heat dissipation adjustment sequence allows the two-phase cold plate module and the spray module to act precisely on different heat dissipation adjustment areas, avoiding the energy waste caused by the simultaneous operation of both modules across the entire area, and further improving the operating economy and flexibility of the heat dissipation system. Attached Figure Description

[0056] Figure 1 This is a timing diagram of the intelligent heat dissipation system for charging piles based on two-phase cold plates and spray systems as described in this invention.

[0057] Figure 2A flowchart illustrating the operation of the heat dissipation mode analysis unit;

[0058] Figure 3 A flowchart for heat source analysis in the heat dissipation mode analysis unit;

[0059] Figure 4 A flowchart for defining the unit's operation in the environmental heat dissipation zone;

[0060] Figure 5 A flowchart illustrating how the execution unit works. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see Figure 1 This invention provides an intelligent heat dissipation system for charging piles based on a two-phase cold plate and a spray system, the system comprising:

[0063] The internal space of the charging pile is divided into multiple independent heat dissipation adjustment zones based on its geometric structure and heat transfer characteristics by a heat dissipation zone division unit. A data acquisition unit collects real-time temperature distribution data of the contact area using a high-precision infrared temperature sensor array deployed on the surface of the charging pile's outer shell, and transmits this data to the central processing unit in a two-dimensional matrix format. After receiving the temperature distribution data, the heat dissipation pattern analysis unit identifies the target heat dissipation area using a pattern recognition algorithm, and further divides the heat dissipation adjustment zone covered by this area into a primary heat dissipation zone and an auxiliary heat dissipation zone. An environmental heat dissipation zone determination unit, based on the principle of heat diffusion, selects environmental heat dissipation zones that need to participate in coordinated heat dissipation from the heat dissipation adjustment zones not covered by the target heat dissipation area. A heat dissipation weight construction unit calculates the difference in thermodynamic parameters between any two adjacent heat dissipation adjustment zones in the heat dissipation treatment set and constructs a heat dissipation weight graph in graph form. A heat dissipation sequence generation unit processes the heat dissipation weight graph using a path optimization algorithm to generate the optimal heat dissipation adjustment sequence. A drive execution unit parses this sequence and controls the phase change refrigeration cycle of the two-phase cold plate module and the liquid cooling system of the spray module to perform zoned heat dissipation operations.

[0064] Example 1: See Figure 2In the actual operation of the intelligent heat dissipation system for charging piles, the heat dissipation mode analysis unit constructs a three-dimensional temperature distribution matrix on the surface of the charging pile's outer shell through a contact area identification module. This matrix consists of 256 high-precision infrared temperature sensors, which are evenly distributed in an 8×32 array on the inner surface of the charging pile's outer shell, with a sampling frequency of 10Hz. The temperature data recorded by each sensor node includes three-dimensional coordinate information (X-axis position, Y-axis position, and temperature value), forming a real-time temperature matrix with dimensions of 8×32×1. During the system initialization phase, 72 hours of historical temperature data are continuously collected and stored as a time-series temperature matrix set.

[0065] The contact area identification module uses the K-nearest neighbor algorithm to process the three-dimensional temperature distribution matrix. The algorithm sets the neighborhood radius parameter to 3 sensor spacing units and the temperature difference threshold to ±2.5℃. For each sensor node, the temperature difference between it and its eight adjacent nodes is calculated. When the temperature difference between a node and more than half of its adjacent nodes remains below the threshold, that node is marked as a core temperature point. All interconnected core temperature points are automatically aggregated into temperature feature similarity regions, generating a set of temperature feature similarity regions containing 5-8 independent regions.

[0066] In the region boundary identification stage, the internal temperature gradient of each region with similar temperature characteristics is calculated. The boundary line is scanned point by point, and the rate of temperature change between the boundary point and its external neighbor is calculated. When the rate of temperature change of a boundary segment continuously exceeds 3℃ / cm and remains so for more than 5 seconds, that boundary segment is marked as a valid contact boundary. Finally, a set of valid contact regions containing 3-5 valid boundaries is selected, and their boundary coordinates are stored in a linked list structure in the standard contact region model database. When the region matching module is working, the real-time temperature matrix is ​​first normalized. The current matrix is ​​compared with 20 sets of template matrices in the standard model database at multiple scales: the first stage uses a 16×16 macroscopic window to scan the overall temperature distribution and calculate the Pearson correlation coefficient; the second stage uses a 5×5 microscopic window to focus on the boundary region and calculate the local temperature distribution similarity. When the macroscopic correlation coefficient exceeds 0.85 and the microscopic similarity is higher than 0.7, it is considered a valid match. After a successful match, abnormal temperature rise regions in the current temperature matrix are extracted, and their boundary coordinates are smoothed using a B-spline curve fitting algorithm to form a sequence of boundary coordinates for the target heat dissipation region.

[0067] When the area mapping module performs coordinate transformation, it calls a preset spatial mapping parameter table. This parameter table contains the correspondence between the physical coordinates of 256 sensor nodes and the heat dissipation adjustment zones. A three-dimensional coordinate system is established with the lower left corner of the charging pile bottom surface as the origin. In the X-axis direction, every 4 sensor nodes correspond to one heat dissipation adjustment zone (8 zones in total), and in the Y-axis direction, every 8 sensor nodes correspond to one heat dissipation adjustment zone (4 zones in total). When the boundary coordinates of the target heat dissipation area are input, the module executes the following mapping process: First, the boundary coordinates are converted into the nearest sensor node number, and then the corresponding heat dissipation adjustment zone number is retrieved based on the node number. For boundary coordinates that span multiple heat dissipation adjustment zones, the area proportion method is used for classification determination—the projected area of ​​the boundary polygon in each heat dissipation adjustment zone is calculated, and heat dissipation adjustment zones with an area proportion exceeding 15% are included in the target area.

[0068] In a real-world example of a charging station operation, when a localized temperature rise is detected in the charging gun interface area: the contact area identification module uses K-nearest neighbor clustering to identify a high-temperature cluster (average temperature 58.3℃) formed by 6×4 sensors near the charging gun interface, with a temperature difference of 7.2℃ from the surrounding area. The area matching module successfully matches this area with the "gun nozzle load" template in the standard model (macroscopic correlation coefficient 0.91), achieving a boundary identification accuracy of ±2mm. The area mapping module maps this area to four heat dissipation adjustment zones numbered A3, A4, B3, and B4, with zone A3 containing 63% of the high-temperature area and thus identified as the core mapping zone. The final output target heat dissipation area boundary data contains 24 key coordinate points, accurately covering a 15cm×10cm rectangular area around the charging gun interface.

[0069] During continuous system operation, the contact area identification module updates the three-dimensional temperature matrix every 30 seconds. When a newly emerging area with similar temperature characteristics is detected: if the area persists for more than 3 sampling periods (30 seconds) and the temperature gradient change rate exceeds 0.5℃ / s, the standard model update process is automatically triggered. The update process retains 80% of the basic data of the original model, and the data of the newly added area is incorporated into the model library after 3 verifications to ensure that the system adapts to the aging and operating condition changes of the charging pile.

[0070] Example 2: See Figure 3During the operation of the charging pile, the heat source analysis module of the heat dissipation mode analysis unit receives real-time thermal imaging data from an infrared thermal imager. This imager acquires the temperature distribution on the surface of the charging pile's outer casing at a frequency of 5 frames per second, generating a grayscale image matrix with a resolution of 256×256 pixels. The grayscale value of each pixel corresponds to a specific temperature, and the conversion relationship follows a pre-calibrated temperature-grayscale curve. When the charging pile is in a high-power charging state, the heat source analysis module initiates a feature extraction process: first, the original thermal image is preprocessed, using a bilateral filtering algorithm to eliminate sensor noise while retaining edge information of temperature abrupt changes. Then, connected component analysis technology is used to identify high-temperature regions in the thermal image, setting a grayscale threshold of 180 (corresponding to 65℃) as the region segmentation benchmark, and marking all interconnected pixel blocks exceeding the threshold.

[0071] In the heat source feature extraction stage, morphological features are calculated for each high-temperature pixel block. The equivalent ellipse major axis of the block is extracted as the main direction of heat diffusion. The standard deviation of pixel grayscale values ​​within the block is calculated as a temperature uniformity index, and the ratio of the block's perimeter to its area is calculated as an edge sharpness coefficient. For typical hot spots appearing in the charging gun interface area, the characteristic data are as follows: the major axis direction is consistent with the charging gun insertion angle (approximately 15° tilt), the temperature uniformity index is in the range of 0.18-0.25, and the edge sharpness coefficient remains around 0.35. These feature parameters are stored in a structured data format in the thermal feature database.

[0072] When the dominant heat dissipation zone determination module is working, it calls the material parameter table of the heat dissipation adjustment zone. This parameter table records the thermal conductivity of the shell material of each heat dissipation adjustment zone. For example, the charging gun interface area uses aluminum alloy (thermal conductivity 117 W / (m·K)), while the control cabinet area uses engineering plastic (thermal conductivity 0.2 W / (m·K)). When the module calculates the heat flux density value of each heat dissipation adjustment zone, it combines real-time temperature gradient data: 9 temperature measurement points are arranged in the A3 heat dissipation adjustment zone (the core area of ​​the charging gun interface), and the X / Y bidirectional temperature gradient is calculated based on the point spacing and temperature difference. The heat flux density calculation adopts Fourier's law, taking the product of the thermal conductivity of aluminum alloy and the temperature gradient. When the heat flux density in the southeast quadrant of the A3 zone is detected to continuously exceed 8 kW / m² for more than 10 seconds, this area is marked as the dominant heat dissipation zone, and the marking signal triggers the boost command of the two-phase cold plate module.

[0073] The auxiliary region determination module, based on thermal conduction gradient analysis, sets up virtual detection paths along the main direction of heat diffusion, starting from the boundary of the dominant heat dissipation zone. Three radial detection lines are arranged in the B3 heat dissipation adjustment zone (adjacent to zone A3), with 10 equidistant virtual temperature measurement points on each line. The theoretical temperature value of each point is calculated using the thermal conduction equation, and compared with the actual measured value to determine the thermal conduction efficiency. When the thermal conduction gradient (temperature decay rate per unit distance) of the western region of zone B3 remains within the range of 0.15-0.2℃ / cm, and the temperature difference with the dominant heat dissipation zone is less than 12℃, this region is marked as an auxiliary heat dissipation zone. A region growth algorithm is used during the marking process: using the boundary pixels of the dominant heat dissipation zone as seed points, the region expands towards adjacent heat dissipation adjustment zones, stopping growth when the heat flux density of the expanded region drops to 30%-60% of that of the dominant zone.

[0074] In a real-world operation under high summer temperatures, after three consecutive fast charging operations, the heat source analysis module detected an elliptical hot spot (maximum temperature 71.3℃) of 98×65 pixels in the charging gun interface area, with its major axis coinciding with the charging gun axis. The dominant region determination module calculated the heat flux density distribution in area A3 and found a peak area of ​​12.3kW / m² in the southeast quadrant, immediately marking this quadrant as the dominant heat dissipation area. The auxiliary region determination module tracked the heat diffusion path and found that heat was conducted along the metal components of the charging gun bracket to area B3, forming a transition area with a gradient of 0.18℃ / cm on the west side of area B3, which was then marked as the auxiliary heat dissipation area. The system drive unit responded to the marking results by activating the two-phase cold plate secondary cooling mode in area A3 and simultaneously activating the directional atomization cooling of the spray module in area B3.

[0075] During continuous system monitoring, the determination of the auxiliary heat dissipation zone exhibits dynamic adjustment characteristics. After the charging pile finishes charging, the temperature of the primary heat dissipation zone decreases at a rate of 1.2℃ / s, and the auxiliary zone determination module updates the heat conduction gradient data in real time. If the temperature decrease rate of the auxiliary heat dissipation zone lags behind the primary zone by more than 15%, the system automatically increases the spray intensity in that area; conversely, when the temperature of the auxiliary zone drops to the ambient temperature +5℃ threshold, the auxiliary marking is automatically removed. This dynamic marking mechanism effectively avoids energy waste caused by excessive cooling in the intermittent operation mode of the charging pile.

[0076] For handling special operating conditions, when multiple heat dissipation spots are detected: the system runs multiple decision threads in parallel. In a dual-gun simultaneous charging scenario, the heat source analysis module identified two independent heat spots (located in regions A3 and C2, respectively). The dominant region determination module establishes an independent heat flux density distribution map for each heat spot, marking the two dominant heat dissipation regions. The auxiliary region determination module calculates the heat diffusion path of its respective dominant region. When it detects that the two auxiliary regions overlap in region D1, it initiates a conflict resolution protocol—recalculating the comprehensive thermal gradient of the overlapping region based on the principle of heat flux density superposition, marking this region as a dual auxiliary heat dissipation region, and triggering a collaborative heat dissipation strategy.

[0077] Example 3: See Figure 4 The environmental heat dissipation zone determination unit establishes a spatial location map of the heat dissipation adjustment zones during system operation. This map is represented by a graph structure, where vertices correspond to the geometric center coordinates of 36 heat dissipation adjustment zones, and edges represent the connectivity between adjacent zones. Each vertex attribute includes a zone number, 3D coordinates, and material code; each edge attribute includes connectivity type (directly adjacent / indirectly adjacent), heat transfer coefficient, and distance weight. The Delaunay triangulation algorithm is used during map construction to ensure the accuracy of spatial relationships, forming a topological network with 78 connected edges.

[0078] When using the density peak clustering algorithm to identify the core influencing region of the dominant heat dissipation zone, it first calculates the local density index of each heat dissipation regulation zone. For any region... Its local density Defined by the following formula:

[0079]

[0080] in: Indicates the area and Euclidean distance, The cutoff distance parameter is set to 1.5 times the average interval between regions. The relative distance between each region is also calculated. This refers to the minimum distance from the local density region to any region with a higher density. Regions whose local density and relative distance product exceeds a set threshold are selected as seed points for the core influence region. In the actual operation of the charging pile, when region A3 is marked as the dominant heat dissipation area, the algorithm calculates its local density to be 3.28 and its relative distance to be 2.1 (normalized value), which is significantly higher than other regions, and therefore it is determined as the core seed point.

[0081] During the regional growth stage, expansion begins from the seed point and proceeds along the edges of the spatial relationship map. The expansion condition considers the heat transfer efficiency factor. This factor is determined by the material combination between regions: 0.8 for metal-metal connections, 0.3 for metal-plastic connections, and 0.1 for plastic-plastic connections. Growth stops when the product of the cumulative heat transfer efficiency of the extended path is less than 0.15. The resulting core influence region contains 6-8 heat dissipation regulation zones, forming a thermal influence cluster centered on the dominant heat dissipation zone.

[0082] When using the thermal diffusion attenuation model to determine the extent of the environmental heat dissipation zone, an unsteady-state heat conduction equation is established:

[0083]

[0084] in: Let be the temperature distribution function. For time variables, The thermal diffusivity of the material, The surface heat dissipation coefficient, Let be the ambient temperature. The equation is solved using the finite difference method to predict the temperature field evolution over the next 180 seconds. Within the non-target heat dissipation region, areas with a predicted temperature rise rate exceeding 0.4℃ / min are selected as candidate ranges for the ambient heat dissipation region.

[0085] The parameter calculation module of the heat dissipation weight construction unit obtains the heat load parameters of each heat dissipation adjustment zone in real time. Heat load parameters From the current temperature Rated temperature and heat capacity Joint decision: Heat dissipation efficiency parameters Then, the surface area of ​​the radiator is taken into account. Surface emissivity and cooling medium flow rate : ,in and This is a weighting factor, adjusted according to the type of heat sink.

[0086] The difference calculation module uses a fuzzy logic system to process parameter differences between adjacent regions. The system input is the difference in thermal load parameters. Difference between heat dissipation efficiency parameter The output is the difference in heat dissipation regulation. The fuzzing stage will It is divided into three fuzzy sets: "low", "medium" and "high", with a threshold of [200, 500] J / ℃; The data is divided into three fuzzy sets: "small," "medium," and "large," with a threshold value of [0.3, 0.6]. The fuzzy rule base contains 9 rules, such as: "If..." High and Large, then "Maximum". Defuzzification uses the centroid method to calculate the precise output value, and finally obtains the normalized difference in the range of 0-1.

[0087] In a practical operation example, when the charging pile performs fast charging: the environmental heat dissipation zone determination unit detected a temperature rise rate of 0.47℃ / min in zone B2 (near the charging gun interface but not the target area), and included it in the environmental heat dissipation zone. The parameter calculation module measured the heat load parameter of zone B2 to be 385J / ℃, and the heat load parameter of the adjacent zone A3 to be 820J / ℃, with heat dissipation efficiency parameters of 0.55 and 0.72, respectively. The difference calculation module calculated... J / ℃ (moderately high). (Smaller), output the difference degree according to the fuzzy rules. The difference value is added to the corresponding edge of the heat dissipation weight graph, affecting the generation of subsequent heat dissipation sequences. The system updates the determination result of the environmental heat dissipation zone every 5 minutes. When the temperature of the dominant heat dissipation zone decreases or the charging power changes, the range of the core influence area is adjusted accordingly, and the division of the environmental heat dissipation zone is dynamically updated. This dynamic adjustment mechanism enables the system to adapt to the heat dissipation requirements of the charging pile under different operating conditions.

[0088] Example 4: See Figure 5 The heat dissipation sequence generation unit receives the heat dissipation weight map and converts it into a 36×36 adjacency matrix. The row and column indices of this matrix correspond to the heat dissipation adjustment zone numbers, and the matrix element values ​​store the difference in heat dissipation adjustment between regions. For non-adjacent regions, the matrix element values ​​are set to infinity. A sparse matrix storage technique is used during the conversion process, retaining only 128 valid connections to reduce computational complexity. When solving for the optimal path sequence based on the Hungarian algorithm, a cost matrix is ​​first constructed. The reciprocal of each element in the adjacency matrix is ​​taken, so that the smaller the difference, the smaller the cost. The algorithm performs row reduction: subtracting the minimum value of each row element ensures that each row contains at least one zero element. Then, column reduction is performed: subtracting the minimum value of each column element ensures that each column contains at least one zero element. By drawing lines to cover all zero elements, the optimal solution is found when the minimum number of lines equals the matrix order. In actual operation, this algorithm can complete the solution of a 36-order matrix within 200 milliseconds and output the optimal access sequence containing 36 regions.

[0089] When optimizing the sequence based on operational constraints, the system considers three types of constraints: minimum heatsink running time constraint (the cold plate module must run for at least 30 seconds each time it starts up), power gradient constraint (power variation between adjacent areas must not exceed 40% of the rated value), and equipment recovery time constraint (the interval between two heat dissipation operations in the same area must be at least 60 seconds). The optimization process uses a constraint propagation algorithm, which first detects node pairs that violate constraints in the initial sequence, and then adjusts them by swapping nodes and inserting no-operations. In one optimization process, it was detected that the power difference between region 12 and region 13 in the sequence reached 52%, violating the power gradient constraint. By inserting a no-operation delay after region 12, the power variation was reduced to 38%.

[0090] The timing control module of the drive execution unit parses the optimized heat dissipation adjustment sequence, which is stored in XML format and includes parameters such as the start timestamp, duration, cooling power level, and spray intensity level for each heat dissipation adjustment zone. The module generates an execution schedule accurate to milliseconds based on the system clock and triggers the heat dissipation operation of each zone via a hardware timer. The power prediction module uses a long short-term memory network model to process historical heat dissipation power data. The network input is the temperature, power, and ambient humidity time series of the past 30 minutes (sampling interval 10 seconds), and the output is the power prediction value for the next 5 minutes. The network structure contains 3 hidden layers, each with 64 neurons, using the tanh activation function. The training data comes from three months of operation records of the charging pile, including heat dissipation power patterns under different seasons and charging powers. When it is predicted that the two-phase cold plate module needs phase change triggering, the output phase change trigger parameters include: refrigerant flow setpoint, evaporation pressure target value, and superheat control range; the flow control parameters for the spray module include: water pump speed, nozzle opening, and atomized particle size distribution.

[0091] The collaborative drive module executes control commands based on predicted parameters. The two-phase cold plate module uses a PID controller to regulate the refrigerant flow rate, aiming to control the evaporator superheat within the range of 4-6℃. The spray module controls the opening and closing frequency of the solenoid valve through pulse width modulation to maintain the spray intensity within the range of 0.8-1.2 L / min·m². The module achieves synchronous control of two heat dissipation modes: when a sharp rise in the area temperature is detected, the cold plate module first activates phase change cooling, and the spray module begins auxiliary heat dissipation 300 milliseconds later; when the temperature tends to stabilize, the spray module gradually reduces its intensity, while the cold plate module maintains its basic cooling power. Actual operating data is recorded in Table 1.

[0092] Table 1: Execution Record of Heat Dissipation Sequence

[0093]

[0094] During execution, the timing control module monitors the completion status of each node. If the heat dissipation effect in a certain area fails to meet the expected target (temperature drop rate less than 70% of the set value), the module automatically extends the duration of that area by up to 10 seconds and adjusts the timestamps of subsequent sequences. Simultaneously, the power prediction module updates its prediction parameters every 30 seconds, adjusting refrigerant flow and spray intensity based on real-time temperature changes. This dynamic adjustment mechanism ensures that the heat dissipation operation always matches the actual heat load. The collaborative drive module also handles special operating conditions: when the ambient humidity exceeds 85%, it automatically reduces the spray intensity by 20% to prevent condensation; when abnormal refrigerant pressure is detected, a safety mode is activated, limiting the cold plate power to within 60% of the rated value. All control parameters and operating status are written to the system log in real time for subsequent analysis and optimization.

[0095] Example 5: The abnormal heat dissipation detection unit continuously monitors the temperature change rate of the heat dissipation processing unit, collecting data through 144 temperature sensors deployed in 36 heat dissipation regulation zones. The sensors collect temperature values ​​at 4-second intervals, and the system calculates the instantaneous temperature change rate for each heat dissipation regulation zone, which is the difference between the current temperature value and the temperature value four seconds ago divided by the time interval. This change rate data is stored in time series format, with each heat dissipation regulation zone independently recording the change rate data of its most recent 300 sampling points, forming a circular buffer of length 300. When the temperature change rate of a heat dissipation regulation zone exceeds a set threshold (positive threshold +0.8℃ / s, negative threshold -1.2℃ / s) for three consecutive sampling points, the system triggers the abnormal detection process.

[0096] When constructing the time-series feature vector of the temperature change rate, the pattern matching unit extracts data from the 30 most recent sampling points from the circular buffer. The feature vector contains 10 dimensions: the first 5 dimensions are statistical features of the change rate (mean, variance, skewness, kurtosis, range), and the last 5 dimensions are time-domain features (zero-crossing rate, autocorrelation coefficient, trend slope, fluctuation frequency, number of abrupt change points). These features are normalized to form a standardized 30-dimensional feature vector. The system maintains an abnormal heat dissipation pattern library containing 12 predefined abnormal pattern templates, each corresponding to a typical heat dissipation system failure, such as radiator blockage, refrigerant leakage, spray nozzle blockage, sensor failure, etc.

[0097] A dynamic time warping algorithm is used to calculate the similarity between real-time feature vectors and anomalous pattern templates. The algorithm first constructs a cumulative distance matrix, where rows correspond to the time points of the real-time feature vectors and columns correspond to the time points of the template feature vectors. An optimal curved path is found that minimizes the cumulative distance, allowing for non-linear scaling of the time axis. The similarity score is calculated based on the final cumulative distance, ranging from 0 to 1, with higher values ​​indicating greater similarity. The system sets a similarity threshold of 0.75; when the similarity of a template exceeds this threshold, a successful match is considered achieved.

[0098] When determining the anomaly pattern type based on the similarity ranking results, the system selects the top three highest-scoring templates for comprehensive judgment. If the score of the highest-scoring template exceeds that of the second-highest-scoring template by more than 0.15, the anomaly type corresponding to the highest-scoring template is directly adopted; if the score difference of the top three templates is less than 0.1, a secondary discrimination mechanism is activated, which further considers historical fault records and current operating environment parameters, and determines the final anomaly type through weighted voting.

[0099] The parameter correction unit updates the edge weights of the heat dissipation weight graph upon successful matching. The update process is based on the influence factors of anomaly modes. Each anomaly mode corresponds to an influence coefficient matrix, which defines the degree of impact of that type of anomaly on the heat transfer efficiency between various heat dissipation adjustment zones. For example, the influence coefficient for radiator blockage anomaly is 0.6-0.8, indicating a 20%-40% reduction in heat transfer efficiency; while the influence coefficient for sensor failure anomaly is 1.2-1.5, indicating a need to strengthen heat dissipation adjustment. The system selects the corresponding influence coefficient matrix based on the determined anomaly type and adjusts the weights of all edges in the heat dissipation weight graph accordingly.

[0100] In a real-world operating example, the system detected an abnormal temperature change rate in the B3 heat dissipation regulation zone: the rate of change remained between +0.9℃ / s and +1.1℃ / s for five consecutive sampling points, significantly exceeding the positive threshold. The pattern matching unit extracted the feature vector of this region and calculated its similarity to the abnormal pattern library. The dynamic time warping algorithm showed a similarity of 0.83 with the "partial radiator blockage" template, 0.71 with the "insufficient refrigerant flow" template, and 0.62 with the "sudden change in ambient temperature" template. Based on the score differences, the system determined it to be an abnormality of partial radiator blockage.

[0101] The parameter correction unit calls the influence coefficient matrix corresponding to the heat sink blockage. This matrix defines the influence coefficient of the blocked area on the heat transfer of adjacent areas as 0.7, and the influence coefficient on areas separated by one region as 0.9. The system updates the weight values ​​of all edges connected to area B3 in the heat dissipation weight graph: the weight of the edge between B3 and A3 is multiplied by 0.7, the weight of the edge between B3 and C3 is multiplied by 0.7, the weight of the edge between B3 and B2 is multiplied by 0.7, and the weight of the edge between B3 and B4 is multiplied by 0.7. At the same time, the edge weights of indirectly connected areas are updated, such as the weight of the edge between A3 and C3, which is multiplied by 0.9. These adjustments cause the heat dissipation sequence generation unit to prioritize the processing of area B3 and its surrounding areas in subsequent calculations.

[0102] The system continuously monitors the effectiveness of anomaly handling, reassessing the anomaly status every 2 minutes. When the temperature change rate in zone B3 returns to the normal range (-0.3℃ / s to +0.4℃ / s) and remains there for 1 minute, the anomaly status is automatically lifted, and the edge weights of the heat dissipation weight graph are gradually restored to normal values. The restoration process employs a gradual adjustment, changing the weight difference by 20% every 30 seconds to avoid sudden changes in system parameters that could lead to instability. All anomaly events and their handling are recorded in the system log, used to optimize the anomaly pattern library and impact coefficient matrix.

[0103] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0104] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A two-phase cold plate and spray-based charging pile intelligent heat dissipation system, characterized in that, The system comprises: A heat dissipation region division unit is configured to divide the inside of the charging pile into a plurality of heat dissipation adjustment regions; A data acquisition unit is configured to acquire temperature distribution data of a contact region on the surface of the charging pile shell in real time; A heat dissipation mode analysis unit is configured to identify a target heat dissipation region according to the temperature distribution data of the contact region, and divide the heat dissipation adjustment region covered by the target heat dissipation region into a dominant heat dissipation region and an auxiliary heat dissipation region; An environmental heat dissipation region determination unit is configured to determine an environmental heat dissipation region in the heat dissipation adjustment region not covered by the target heat dissipation region, and the dominant heat dissipation region, the auxiliary heat dissipation region and the environmental heat dissipation region constitute a heat dissipation processing set; A heat dissipation weight construction unit is configured to calculate a heat dissipation adjustment difference between any two adjacent heat dissipation adjustment regions in the heat dissipation processing set, and construct a heat dissipation weight graph with the heat dissipation adjustment regions as vertices and the heat dissipation adjustment differences as edges; A heat dissipation sequence generation unit is configured to generate a heat dissipation adjustment sequence based on the heat dissipation weight graph; A driving execution unit is configured to drive the two-phase cold plate module and the spraying module to perform a heat dissipation operation according to the heat dissipation adjustment sequence.

2. The two-phase cold plate and spray-based charging pile intelligent heat dissipation system according to claim 1, characterized in that, The heat dissipation mode analysis unit comprises: A contact region identification module is configured to establish a standard contact region model according to historical temperature data; A region matching module is configured to match the temperature distribution data of the contact region acquired in real time with the standard contact region model, and identify the boundary of the target heat dissipation region under the current operating state of the charging pile; A region mapping module is configured to map the boundary of the target heat dissipation region to the spatial coordinates of the heat dissipation adjustment region.

3. The two-phase cold plate and spray-based charging pile intelligent heat dissipation system according to claim 2, characterized in that, The contact region identification module is specifically configured to: Construct a three-dimensional temperature distribution matrix of the surface of the charging pile shell; Perform clustering analysis on the three-dimensional temperature distribution matrix by using a K-nearest neighbor algorithm to generate a temperature feature similar region set; Screen effective contact regions in the temperature feature similar region set according to temperature variation gradients.

4. The two-phase cold plate and spray-based charging pile intelligent heat dissipation system according to claim 1, characterized in that, The heat dissipation mode analysis unit further comprises: A heat source analysis module is configured to extract heat imaging feature data in the target heat dissipation region; A dominant region determination module is configured to mark the heat dissipation adjustment region corresponding to the heat imaging feature data as a dominant heat dissipation region according to a heat flow density threshold; An auxiliary region determination module is configured to mark the heat dissipation adjustment region adjacent to the dominant heat dissipation region as an auxiliary heat dissipation region according to a heat conduction gradient.

5. The two-phase cold plate and spray-based charging station intelligent heat dissipation system of claim 1, wherein, The environmental heat dissipation region determination unit is specifically configured to: Establish a spatial position relationship graph of the heat dissipation adjustment region; Identify a core influence region of the dominant heat dissipation region by using a density peak clustering algorithm; Determine the range of the environmental heat dissipation region in the non-target heat dissipation region according to a heat diffusion attenuation model.

6. The two-phase cold plate and spray-based charging station intelligent heat dissipation system of claim 1, wherein The heat dissipation weight construction unit comprises: A parameter calculation module is configured to obtain heat load parameters and heat dissipation efficiency parameters of each heat dissipation adjustment region; A difference calculation module is configured to process the heat load parameter difference and the heat dissipation efficiency parameter difference of adjacent heat dissipation adjustment regions by using a fuzzy logic algorithm, and output a heat dissipation adjustment difference.

7. The two-phase cold plate and spray-based charging pile intelligent heat dissipation system according to claim 6, characterized in that, The heat dissipation sequence generation unit is specifically configured to: Convert the heat dissipation weight graph into an adjacency matrix; Solve an optimal path sequence of the adjacency matrix based on the Hungarian algorithm; Optimize the optimal path sequence according to the working state constraint condition of the heat dissipation adjustment region.

8. The two-phase cold plate and spray-based charging station intelligent heat dissipation system of claim 1, wherein, The driving execution unit comprises: A timing control module is configured to analyze execution node parameters in the heat dissipation adjustment sequence; A power prediction module is configured to process historical heat dissipation power data by using a long short-term memory network model to generate phase change triggering parameters of the two-phase cold plate module and flow control parameters of the spraying module; A cooperative driving module is configured to synchronously adjust the refrigerant flow rate and the spraying intensity according to the phase change triggering parameters and the flow control parameters.

9. The two-phase cold plate and spray-based charging station intelligent heat dissipation system of claim 1, wherein, The system further comprises: An abnormal heat dissipation detection unit is configured to monitor temperature change rates of the heat dissipation processing set; A pattern matching unit is configured to compare a real-time temperature change rate curve with a preset abnormal heat dissipation pattern library; A parameter correction unit is configured to update edge weight values of the heat dissipation weight graph when a matching degree exceeds a set threshold.

10. The two-phase cold plate and spray-based charging pile intelligent heat dissipation system according to claim 9, characterized in that, The pattern matching unit is specifically configured to: construct a time series feature vector of the temperature change rate; calculate a similarity degree between the time series feature vector and an abnormal heat dissipation pattern template by using a dynamic time warping algorithm; and determine an abnormal heat dissipation pattern type with the highest matching degree according to a similarity degree sorting result.

Images